While running gromacs simulations why domain decomposition appears.

i could be able to minimize the system using -nt 1. But still the same problem appears in further nvt and npt ensembles.

Initially i got error

"some parameters defined multiple times"

followed by changing coulumb and vdw cotoff from 1.0 to 1.4 error was

" here is no domain decomposition for 48 ranks that is compatible with the given box and a minimum cell size of 7.03502 nm

Change the number of ranks or mdrun option -rdd or -dds"

.

in this simulation am applying distance restraints for three coupled metal atoms & its coordinating atoms with in this distance of 4 angstrom.

i tried mdrun with different -dlb, -npme, lesser -rdd and '-dds yes' options. It decreases the DD cell size but still error is same.

How i can solve this error? How dd will be varied for non-homogeneous systems. I have included the screenshots of PDB before and after minimisation, log and parameter files.